Eight herpesviruses cause disease in humans. Most are virtually ubiquitous, and collectively they impose a huge burden on human health, ranging from the pain and inconvenience of genital herpes caused by herpes simplex viruses (HSV) to the devastating congenital disease and life-threatening pneumonia caused by human cytomegalovirus to the cancers associated with Epstein-Barr virus (EBV). For some there are safe and effective antiviral drugs, but for others the drugs that might be effective have serious side effects that rule them out for many applications. Clearly there is much more to be done to reduce the vast impact of herpesviruses. This proposal focuses on EBV, not because it is the herpesvirus most in need of antiviral treatments but because it is the most amenable to studying membrane fusion, an essential process for entry of all herpesviruses into cells. There are two reasons for this. (i) In the context of epithelial cells, only glycoproteins gB and gH/gL are needed for fusion, with gH/gL believed to bind gB and trigger a conformation change that activates its fusion activity. For most other herpesviruses, gH/gL must itself be activated by an additional viral protein before it can activate gB. (ii) EBV gB can also be conformationally activated by heat, providing an experimental mechanism for distinguishing between gB mutations that prevent gH/gL binding and those that prevent the conformation change. Binding of gH/gL to gB is a potential target for antiviral drugs, but the rational design of such drugs is currently not possible because the 3D structure of the binding site is unknown - neither the gB residues that constitute that site nor the structure of gB before activation (pre-fusion gB, to which gH/gL most likely binds) have been determined. The known structures are for post-fusion gB (the structure after membrane fusion), and the pre-fusion form has resisted crystallization. The experiments proposed will identify the residues that form the binding site and identify mutations that lock gB in its pre-fusion state, thus enabling the future determination of that structure and subsequently rational drug design for preventing EBV entry into cells. Most importantly, this proof-of-principle will hasten the rational design of antiviral drugs for any herpesvirus. The aims are therefore: (1) to identify the gB residues required for binding of gH/gL; (2) to identify mutations that prevent gB from undergoing the conformational change from its pre-fusion form to its post-fusion form. A novel approach to selection of sites for mutagenesis has been developed. Activation of correctly folded mutants by gH/gL, by heat, or by neither, combined with susceptibility to proteolysis, will identify those which cannot bind gH/gL, those which can bind gH/gL but not switch to the fusogenic conformation, and those which fold during synthesis into the post-fusion form and therefore cannot be activated. The roadblocks standing in the way of drugs designed to prevent entry of herpesviruses into cells will then have been removed.